Replied: Mon, 20 Jan 1997 18:13:50 -0500
Replied: "Stefan Monnier <monnier-stefan@cs.yale.edu> mills@udel.edu, Stefan Monnier <monnier@cs.yale.edu>"
Received: from copland.udel.edu by huey.udel.edu id aa06776; 18 Jan 97 0:03 EST
Received: from POWERED.ZOO.CS.YALE.EDU (ZOO-GW.CS.YALE.EDU [128.36.0.5]) by copland.udel.edu (8.8.4/8.7.3) with ESMTP id AAA26324 for <mills@udel.edu>; Sat, 18 Jan 1997 00:03:16 -0500 (EST)
Received: from cs.yale.edu by POWERED.ZOO.CS.YALE.EDU (8.7/res.host.cf-3.5)
	with ESMTP id AAA29794; Sat, 18 Jan 1997 00:03:14 -0500
From: Stefan Monnier <monnier-stefan@cs.yale.edu>
To: mills@udel.edu
Cc: Stefan Monnier <monnier@cs.yale.edu>
Subject: xntp3-5.88 on Linux-redhat-4.0
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Sat, 18 Jan 1997 00:03:12 -0500
Message-ID: <15418.853563792@cs.yale.edu>
Sender: monnier-stefan@cs.yale.edu


the util/ntptime.c needed a few corrections to compile:

- line 42: the HAVE___ADJTIMEX seems to determine whether the function
  __adjtimex is available and yet, it uses __adjtimed which leads to a
  link error. It was probably just a typo.

- the pll_trap function has to be protected by "#ifdef SIGSYS" because
  gcc doesn't like the fact that "env" is unknown (even though the function
  will never be used).

wishes:

- support for a shared libntp.so library. Admittedly, it's not too hard
  to do it by hand, which reduces the incentive to improve the makefiles
  and the configuration scripts.

- an sntpd which would be a stripped down xntpd for SNTP (even if my PC
  has 64MB of memory, it seems excessive to devote 588KB (RSS claimed by top
  and ps) of it just to keep my clock synchronised. Do you happen to know
  if such a client exists for unix (the NTP webpage seems to indicate that
  it is not the case, but .... maybe ? ... no ? oh well)


	Stefan


